A plate layout optimization method based on fusion of immune genetic algorithm and tabu search

Abstract

This invention discloses a method for optimizing board layout based on the fusion of immune genetics and tabu search, belonging to the field of intelligent manufacturing and operations research. The method includes: collecting data and initializing a population; performing decoding operations on each chromosome individual in the population; determining whether the current iteration count meets a preset tabu search cycle threshold, and performing local depth fine-tuning using a tabu search module; calculating the Hamming distance between individuals within the population to assess gene similarity, and calculating the immune concentration of each individual; combining fitness and immune concentration to calculate the overall selection probability of the individual, and then generating a new generation of the population; repeating the above steps, and monitoring whether the physical utilization rate of the globally optimal layout scheme meets the evolutionary stagnation condition. This method solves the problems of illegal solutions, severe population homogenization, and easy local optimization deadlock in traditional layout algorithms, effectively improving the overall physical utilization rate of large-scale board layout.

Description

Technical Field

[0001] This invention belongs to the field of intelligent manufacturing and operations optimization technology, specifically involving a plate layout optimization method based on the fusion of immune genetics and taboo search. Background Technology

[0002] In the field of solid wood processing and intelligent manufacturing, multi-specification one-dimensional nesting aims to find the optimal cutting combination to maximize material utilization, which is a typical NP-hard problem. Currently, the industry often uses heuristic algorithms such as genetic algorithms or tabu search for nesting optimization, but there are still many technical bottlenecks that need to be solved when dealing with large-scale cases.

[0003] First, traditional nesting algorithms rely heavily on static first-fit or best-fit rules in their underlying decoding logic. Due to the lack of foresight regarding the subsequent sequence of boards to be nested, they are prone to wasting space in the early stages of nesting and may even produce illegal solutions that violate physical space constraints.

[0004] Secondly, population evolution algorithms, represented by genetic algorithms, are prone to getting trapped in local minima in the later stages of optimization, leading to severe homogenization of the population. The root cause of this phenomenon is that conventional random crossover and mutation operators are prone to destroying the already formed good local sampling combinations when exploring new solutions; while directly introducing local optimization strategies such as tabu search to break the deadlock often leads to a surge in the computational complexity of the hybrid architecture, making it difficult to achieve a balance between optimization depth and computation time.

[0005] Furthermore, existing algorithms generally lack a real-time perception and extraction mechanism for excellent nesting patterns during dynamic evolution. This prevents high-value local nesting features (such as combinations of boards that have achieved extremely high fill rates) from being stably passed down through generations, further limiting the possibility of the overall material utilization rate approaching its physical limits. Therefore, the industry urgently needs a comprehensive nesting method that can avoid illegal solutions from the bottom up, effectively escape local deadlocks, and stably inherit excellent nesting features. Summary of the Invention

[0006] To address the problems existing in the background technology, this invention proposes a board layout optimization method based on the fusion of immune genetics and tabu search. This method solves the problems of illegal solutions, severe population homogenization, and easy getting stuck in local optimization deadlock in traditional layout algorithms, and effectively improves the overall physical utilization rate of large-scale board layout.

[0007] Technical Solution: To solve the above-mentioned technical problems, the present invention adopts the following technical solution:

[0008] A plate layout optimization method based on the fusion of immune genetics and taboo search includes the following steps:

[0009] S1. Collect target board demand data and original board inventory data for solid wood boards, assign serial numbers to the target boards, and initialize the population using a random shuffling strategy to obtain multiple chromosome individuals.

[0010] S2. Perform decoding operation on each chromosome individual in the population, allocate the target board to the original board according to the gene sequence, and calculate the smooth fitness and physical utilization rate of each individual for the allocated layout. At the same time, update the global optimal layout plan record based on the physical utilization rate.

[0011] S3. Determine whether the current iteration count meets the preset tabu search cycle threshold. If it does, select the top E1 elite individuals in the population with the highest physical utilization rate and input them into the tabu search module to perform local depth fine-tuning. Replace the original elite individuals with the fine-tuned individuals. If it does not meet the threshold, go to S4.

[0012] S4. Calculate the Hamming distance between individuals within the population to assess gene similarity, and calculate the immune concentration of each individual based on gene similarity;

[0013] S5. Combine smooth fitness and immune concentration to calculate the overall selection probability of an individual, and perform genetic evolution operations based on immune selection according to the probability to generate a new generation of population.

[0014] S6. Repeat S2 to S5, and monitor in real time whether the physical utilization rate of the global optimal layout scheme meets the evolutionary stagnation condition during the loop iteration. If it does, perform an adaptive immune clearing operation to break the optimization deadlock until the termination condition is met, and output the global optimal physical utilization rate and its corresponding board layout scheme.

[0015] As a preferred embodiment, in S2, the decoding operation is implemented as follows:

[0016] The decoding operation adopts a dynamic optimal adaptation rule, which is as follows: the target board is read sequentially according to the gene sequence, and it is preferentially allocated to the opened original board with the smallest remaining space after allocation; if all the currently opened original boards cannot accommodate the target board, the subsequent target board sequence is calculated in advance, and a new specification original board with the highest expected filling rate is opened from the inventory.

[0017] As a preferred option, in S2, the smooth fitness and physical utilization rate of each individual are calculated, and the implementation content of the globally optimal sorting scheme record is updated based on the physical utilization rate as follows:

[0018] Calculate the physical utilization rate: Extract the total length of all target boards in the current individual layout plan, and the total length of the original boards actually consumed after allocation. Calculate the global physical utilization rate U of the layout plan by the ratio of the two. The calculation formula is as follows:

[0019] ;

[0020] in, This indicates the current physical utilization rate of an individual; Indicates the total number of target sheet materials; Indicates the first The length of the target sheet material; This indicates the total number of original boards used in this individual layout plan; Indicates the first The original material has been opened to its specified length;

[0021] Calculate the smoothness fitness: Sum the target board lengths allocated to each opened original board, and calculate the single-board fill rate for each original board. Then, average the squares of the single-board fill rates for all opened original boards to obtain the individual smoothness fitness. To enhance the fitness weight of compact layout combinations, the calculation formula is as follows:

[0022] ;

[0023] in, This represents the smooth fitness of the current chromosome individual; This indicates the total number of original boards used in this individual layout plan; Indicates allocation to the first The number of target boards in the original board material; Indicates the first The first in the root board The length of the target plate.

[0024] As a preferred option, in S3, the implementation of local depth fine-tuning by the tabu search module is as follows:

[0025] Multiple neighborhood solutions are generated using a two-point position swap strategy. If the physical utilization rate of a neighborhood solution is higher than the historical global best record, the contempt criterion is triggered, and the neighborhood solution is directly accepted. If the contempt criterion is not triggered and the move is not in the taboo list, the neighborhood solution with the highest fitness is accepted, and the move is added to the fixed-length taboo list.

[0026] As a preferred option, the specific implementation details in S4 are as follows:

[0027] S401. Constructing the feature matrix: Transform the population into a matrix of size... The feature matrix;

[0028] in, This represents the total number of individuals in the population. Indicates the total number of target sheet materials; Indicates the first The first individual The target plate number corresponding to each gene locus;

[0029] S402. Calculate Hamming distance: Compare any two individuals in the feature matrix one by one. and By analyzing gene sequences and counting the number of genes with different values ​​at the same gene locus, the Hamming distance can be obtained. The calculation formula is as follows:

[0030] ;

[0031] in, Represents an individual With individuals Hamming distance between them; Indicates characteristic functions;

[0032] S403. Calculate gene similarity: Assess gene similarity between individuals based on the calculated Hamming distance. The calculation formula is as follows:

[0033] ;

[0034] in, Represents an individual With individuals Genetic similarity between them;

[0035] S404. Calculate the individual's immune concentration. The calculation formula is as follows:

[0036] ;

[0037] in, Indicates the first The immune concentration of an individual; This indicates the preset threshold for determining gene similarity. ( ) indicates an indicator function.

[0038] As a preferred option, the specific implementation details in S5 are as follows:

[0039] The overall selection probability adopts an exponential decay mechanism based on concentration suppression, and the specific calculation method is as follows:

[0040] ;

[0041] in, Indicates the first in the population The overall probability of selection for each individual; , Indicates the first The, the The smoothness of fitness of each individual; , Indicates the first The, the The immune concentration of an individual; This represents the preset concentration inhibition coefficient; This represents an exponential function with the natural constant e as its base. It represents the total number of individuals in a population.

[0042] As a preferred option, in S5, the implementation of the genetic evolution operation is as follows:

[0043] The genetic evolution operation is as follows: a gene sequence from a parent chromosome is randomly selected and directly copied to the same position in the offspring. Then, a gene from another parent chromosome that does not appear in the above sequence is filled into the remaining empty space in the offspring in the original order to ensure the absolute conservation of the types and quantities of the target material for the short board.

[0044] As a preferred option, the specific implementation details in S6 are as follows:

[0045] The determination of whether the evolutionary stagnation condition is met and the execution of the adaptive immune purge operation are as follows: when the physical utilization rate of the globally optimal sorting scheme has not improved for M consecutive generations, the algorithm is determined to be trapped in a local extreme value trap and meets the evolutionary stagnation condition, and the adaptive immune purge operation is triggered.

[0046] The adaptive immune clearance operation specifically includes: traversing the current population, identifying individuals with immune concentrations higher than a preset clearance threshold as redundant antibodies and removing them, and then using a random initialization strategy to generate an equal number of new chromosome individuals to fill the population vacancies; and during the clearance process, exemptions are reserved for elite individuals with the highest physical utilization rate (E2) and individuals containing dynamic vaccine sequences.

[0047] Beneficial effects: Compared with the prior art, the present invention has the following advantages:

[0048] (1) The present invention adopts dynamic optimal adaptation rules and look-ahead computation as the underlying decoding strategy, and decentralizes physical space constraints to the rule layer, thereby fundamentally avoiding the generation of illegal solutions and ensuring the absolute legality of recombinant genes.

[0049] (2) This invention introduces a periodic elite taboo fusion architecture. By limiting the scope and frequency of local search, it performs deep fine-tuning on high-fitness elite individuals only when the periodic threshold is met, thus achieving architecture-level computational dimensionality reduction. Combined with the historical global optimal record triggering the contempt criterion, this architecture not only significantly reduces the computational complexity of the hybrid algorithm, but also significantly enhances the algorithm's ability to escape local extrema and approach the theoretical limit of utilization.

[0050] (3) This invention sets up a dynamic vaccine extraction and targeted inoculation mechanism, breaking the limitation of traditional algorithms that rely on static prior knowledge. By extracting the target board combination in the original board with high filling rate as the vaccine sequence in real time during the decoding process, and forcing the sequence to be arranged adjacently and preserved as a whole during the crossover mutation stage, the targeted inheritance of excellent sorting features is realized, effectively preventing the destruction of high-value sorting combinations in the evolution process.

[0051] (4) This invention integrates an adaptive immune clearance mechanism to address premature convergence and optimization deadlock problems. By dynamically adjusting the individual selection probability based on Hamming distance-based concentration suppression, homogeneous individuals are effectively suppressed; and when global evolution stalls, high-concentration redundant individuals are automatically cleaned and random new individuals are injected. Combined with the exemption retention strategy for elites and vaccine sequences, it ensures that the algorithm breaks local deadlock while preventing the loss of core superior genes. Attached Figure Description

[0052] Figure 1 This is the overall flowchart of a plate layout optimization method based on the fusion of immune genetics and taboo search;

[0053] Figure 2 This is a comparison chart of the training process of different sorting algorithms. Detailed Implementation

[0054] The present invention will be further illustrated below with reference to specific embodiments. These embodiments are implemented based on the technical solutions of the present invention, and it should be understood that these embodiments are only used to illustrate the present invention and are not intended to limit the scope of the present invention.

[0055] This embodiment provides a board material layout optimization method based on the fusion of immune genetics and tabu search. It constructs a hybrid optimization architecture combining macro and micro perspectives, mainly consisting of global immune evolution and local tabu search. Global immune evolution is used to explore the layout solution space on a large scale, while local search is used to fine-tune elite layout schemes. This method is based on an immune genetic algorithm framework and is deeply optimized for large-scale board material layout. Its underlying layer uses prospective optimal adaptation rules for decoding and real-time construction of a dynamic vaccine library, supplemented by a forced binding mechanism for vaccine sequences during crossover and recombination, and an adaptive immune clearance mechanism when evolution stagnates. This solves the problems of illegal solutions, severe population homogenization, and easy local optimization deadlock in traditional layout algorithms, effectively improving the overall physical utilization rate of large-scale board material layout.

[0056] like Figure 1 As shown, the specific steps include:

[0057] S1. Collect target board demand data and original board inventory data for solid wood boards, assign serial numbers to the target boards, and initialize the population using a random shuffling strategy to obtain multiple chromosome individuals.

[0058] The system first reads the actual production order data to complete the data collection. The order contains a total of 250 target boards (specific specifications and quantities are: 40 pieces of 1390mm, 35 pieces of 980mm, 35 pieces of 830mm, 35 pieces of 785mm, 35 pieces of 660mm, 35 pieces of 580mm, and 35 pieces of 450mm); and a total of 114 raw boards in stock (specific specifications and quantities are: 22 pieces of 9400mm, 26 pieces of 8200mm, 28 pieces of 7550mm, and 38 pieces of 5500mm).

[0059] Subsequently, the system assigns a globally unique serial number to each of the 250 target boards from 0 to 249. After setting the population size to 60 and the maximum number of iterations to 100, it uses a random shuffling strategy to permutate the serial numbers, generating 60 initial chromosome individuals with a length of 250mm, thus completing the population initialization.

[0060] S2. Perform decoding operation on each chromosome individual in the population, allocate the target board to the original board according to the gene sequence, and calculate the smooth fitness and physical utilization rate of each individual for the allocated layout. At the same time, update the global optimal layout plan record based on the physical utilization rate.

[0061] The decoding operation adopts a dynamic optimal adaptation rule. In specific implementation: when the algorithm reads the target board (e.g., with a length of 1390mm) sequentially according to the gene sequence, it first traverses the currently opened original board and puts it into the opened original board with the smallest remaining space after allocation to achieve the best physical fit; if none of the currently opened original boards can accommodate the target board, the system will not blindly open a new board, but will trigger a forward calculation mechanism to open a new specification original board with the highest expected filling rate from the inventory by forward calculating the subsequent target board sequence.

[0062] In this embodiment, the system simulates placing the target board and its subsequent gene sequence into four unopened original boards of four specifications: 9400mm, 8200mm, 7550mm, and 5500mm, respectively, to calculate the future space utilization potential. Finally, it opens a new original board with the highest expected filling rate from the inventory, thereby avoiding early space waste that is easily caused by traditional algorithms.

[0063] In the decoding operation of step S2, the system calculates the single-piece filling rate of each original board in real time. If it is found that the physical utilization rate of ten target boards [two 1390mm pieces, one 980mm piece, one 830mm piece, two 580mm pieces, and four 450mm pieces] loaded in a certain 7550mm original board reaches 100% (exceeding the preset extreme value threshold of 99.5%), the gene number combination corresponding to these ten boards is immediately extracted as a "vaccine sequence" and stored in the dynamic vaccine library. When generating a new generation population, the vaccine sequence is extracted according to a preset probability (30%), and in the crossover mutation operation, the numbers of these ten target boards are forced to be arranged closely adjacently in the offspring chromosomes, and preserved as an indivisible whole gene block to prevent this extremely excellent local cutting combination from being destroyed by subsequent mutations.

[0064] The implementation details of calculating the fitness and physical utilization rate of each individual, and updating the global optimal sampling plan record based on the physical utilization rate, are as follows:

[0065] (1) Calculate the physical utilization rate: Extract the total length of all target boards in the current individual layout plan, and the total length of the original boards actually consumed after allocation. Calculate the global physical utilization rate U of the layout plan by the ratio of the two. The calculation formula is as follows:

[0066]

[0067] in, This indicates the current physical utilization rate of an individual; Indicates the total number of target sheet materials; Indicates the first The length of the target sheet material; This indicates the total number of original boards used in this individual layout plan; Indicates the first The original material has been opened to its specified length;

[0068] (2) Calculate the smoothness fitness: Calculate the sum of the target board lengths allocated to each opened original board and the single-board fill rate of each original board; then, average the squares of the single-board fill rates of all opened original boards to obtain the individual smoothness fitness. To enhance the fitness weight of compact layout combinations, the calculation formula is as follows:

[0069]

[0070] in, This represents the smooth fitness of the current chromosome individual; This indicates the total number of original boards used in this individual layout plan; Indicates allocation to the first The number of target boards in the original board material; Indicates the first The first in the root board The length of the target plate.

[0071] S3. Determine whether the current iteration number meets the preset taboo search cycle threshold (set to 5 generations). If it does, trigger periodic elite taboo fine-tuning, isolate the very few elite individuals with the best performance in the population, input them separately into the taboo search module to perform local deep mining, and replace the original individuals with the new elites obtained after fine-tuning. If the cycle threshold is not met, proceed directly to the next step (S4).

[0072] The tabu search module performs local depth fine-tuning as follows: For the top E1 (E1=3) elite individuals in terms of physical utilization, a two-point position swap strategy is used to randomly generate 10 neighborhood solutions (i.e., randomly swapping the arrangement order of the two target boards). The generated neighborhood solutions are then evaluated. When making a judgment, the following core logic should be considered:

[0073] First, if the physical utilization rate of a certain neighborhood solution is higher than the historical global best record, that is, if the following formula is satisfied:

[0074]

[0075] This triggers the contempt criterion, which accepts the neighborhood solution directly regardless of any taboo conditions; where, Represents the generated neighborhood solutions; This represents the physical utilization rate of the solution in that neighborhood; This represents the globally optimal physical utilization rate in historical records.

[0076] Secondly, if the contempt criterion is not triggered, and the current move / swap action is not in the fixed-length taboo list T of length 5, then the solution with the highest fitness in the legal neighborhood is accepted, i.e., the solution that satisfies the following formula:

[0077]

[0078] After accepting the solution, the move and exchange action is recorded in the taboo list according to the first-in, first-out principle, guiding elite individuals to conduct in-depth exploration in the local solution space; among which, This represents the smooth fitness of the solution in that neighborhood; This represents the fitness set of all solutions in the currently generated 10 neighborhood solutions that are not in the tabu list T; This represents the operation of finding the maximum value in a set.

[0079] S4. Transform the entire population into a feature matrix, compare the differences in gene arrangement between individuals one by one, and calculate the immune concentration of each individual by evaluating Hamming distance and gene similarity.

[0080] S401. Constructing the feature matrix: Transform the population into a matrix of size... The characteristic matrix of , where, Population size, which is the total number of individuals in the population; The total number of target sheet materials; Indicates the first The first individual The target plate number corresponding to each gene locus;

[0081] S402. Calculate Hamming distance: Compare any two individuals in the feature matrix one by one. and By analyzing gene sequences and counting the number of genes with different values ​​at the same gene locus, the Hamming distance can be obtained. The calculation formula is as follows:

[0082]

[0083] in, Represents an individual With individuals Hamming distance between them; Let x be an indicator function, when x = y, ;otherwise ; Indicates the first The first individual The target plate number corresponding to each gene locus; Indicates the first The first individual The target plate number corresponding to each gene locus; The total number of target sheet pieces.

[0084] S403. Calculate gene similarity: Assess gene similarity between individuals based on the calculated Hamming distance. The calculation formula is as follows:

[0085]

[0086] in, Represents an individual With individuals Genetic similarity between them; Represents an individual With individuals Hamming distance between them; The total number of target sheet pieces.

[0087] S404. Calculate the individual's immune concentration. The calculation formula is as follows:

[0088]

[0089] in, Indicates the first The immune concentration of an individual; The preset threshold for gene similarity determination; ( ) is an indicator function, when the condition is met. The value is 1 when it is true, and 0 otherwise. Represents an individual With individuals Genetic similarity between them; Population size refers to the total number of individuals in the population.

[0090] S5. Combining the smooth fitness and immune concentration in the above dual-track evaluation system, calculate the comprehensive selection probability of each individual, and perform genetic evolution operations based on immune selection according to the probability. The genetic evolution operations specifically include roulette wheel operation, crossover operation and mutation operation, thereby completing immune evolution and generating a new generation of population.

[0091] The overall selection probability adopts an exponential decay mechanism based on concentration suppression, and the specific calculation method is as follows:

[0092]

[0093] in, Indicates the first in the population The overall probability of selection for each individual; , Indicates the first The, the The smoothness of fitness of each individual; , Indicates the first The, the The immune concentration of an individual (i.e., the proportion of individuals whose similarity to other individuals in the population is greater than a threshold). This indicates the preset concentration suppression coefficient (used to adjust the intensity of concentration penalty). It represents an exponential function with the natural constant e as the base; N represents the total number of individuals in the population; Indicates the index number of an individual in the population ( = 1,2, ...,N).

[0094] In practice, the population size N=60, the first... individual fitness The immune concentration calculated using Hamming distance The preset concentration inhibition coefficient is 0.85 (meaning that 85% of the individuals in the population are highly similar to it). Let's set it to 0.8. Using this formula, the probability of this high-concentration individual being selected is... It will be exponentially expressed by the function exp(-0.8). 0.85) This strongly weakens the population, thereby forcibly reducing the reproduction probability of homogeneous individuals and ensuring the diversity of the population's genes.

[0095] In this embodiment, the genetic evolution operation includes a sequential crossover operation, where a continuous gene sequence (e.g., indices 10 to 20) is randomly extracted from parent chromosome A and directly copied to the same position in the offspring chromosome. Subsequently, the remaining genes from another parent chromosome B that do not appear in the extracted sequence are extracted and sequentially filled into the remaining vacancy in the offspring according to the original order of B. This mechanism, while achieving gene recombination, ensures that the type and quantity of the 250 short-board target boards are not lost or duplicated, maintaining the absolute conservation of the actual order materials.

[0096] When generating the new generation population in step S5, vaccine sequences are extracted from the dynamic vaccine library according to a preset probability. When performing crossover and mutation operations, the target plate numbers contained in the extracted vaccine sequence are forced to be arranged adjacently in the offspring chromosomes and preserved as a whole to prevent the excellent sampling combination from being destroyed.

[0097] In this embodiment, the roulette wheel betting operation, crossover operation, and mutation operation are as follows:

[0098] (1) Roulette wheel selection based on immune selection: When performing roulette wheel selection based on immune selection, the system first constructs a probability distribution based on the comprehensive selection probability of each individual, and randomly selects individuals into the crossover pool by simulating roulette wheel rotation. Specifically, the system pre-calculates the cumulative probability of each individual in the population, and the calculation formula is as follows:

[0099]

[0100] in, Indicates the first in the population The cumulative probability of each individual; Indicates the first in the population The overall probability of selection for each individual; And satisfy .

[0101] Subsequently, the system generates a random number r that is uniformly distributed in the interval [0,1]. If this random number satisfies... (Regulation If =0), then the corresponding first Each individual is then selected to enter the parent population. This extraction process will be repeated until the required number of individuals for the set population size are selected.

[0102] At the same time, in order to prevent the accidental loss of superior genes during the selection process, the system will also implement an elite retention strategy, forcibly copying the globally optimal individual with the highest physical utilization rate in the previous generation directly into the new generation population.

[0103] (2) Crossover operation: After the selection of parent individuals is completed, the crossover operation is performed immediately. This embodiment uses a sequential crossover operator to ensure that the types and quantities of target plates remain absolutely conserved during gene recombination.

[0104] In practice, the system first randomly selects two crossover points from the chosen parent chromosome P1, extracts a continuous gene segment between these two points, and directly copies it to the corresponding physical location on the offspring chromosome C1. Then, the system iterates through the other parent chromosome P2 participating in the crossover, removing all duplicate gene sequences already present in the C1 segment, and then filling the remaining gene sequences in P2 into the remaining empty spaces in C1 according to their original order of appearance. Similarly, using P2 as the baseline parent and P1 as an auxiliary feature extraction method, the system symmetrically generates a second offspring chromosome C2. Through this mechanism, the newly generated individuals effectively inherit the excellent crossover features of both parents while fundamentally avoiding the problems of duplicate or omitted plate code that are prone to occur with traditional crossover operators.

[0105] (3) Mutation operation: Finally, in order to further increase the diversity of the population genes and prevent the optimization process from getting stuck in a local deadlock, the system will perform mutation operation on the offspring individuals generated after crossover.

[0106] This embodiment specifically employs a position-exchange mutation operator, according to a preset mutation probability P. m (In this embodiment, the value is set to 0.1) Triggering the mutation mechanism. Once this mechanism is triggered, the system will randomly select two unequal mutation sites, indices m and n, in the current offspring chromosome, and then directly swap the target plate number at index m with the target plate number at index n. The expression for this swap logic is:

[0107]

[0108] in, This represents the original chromosome sequence before the mutation; This represents two randomly generated position indices; This represents the new chromosome sequence generated after the exchange; This indicates a position swap operation, which involves swapping the indices in the sequence. and Swap the elements at that location.

[0109] S6. Repeat steps S2 to S5, and monitor in real time whether the physical utilization rate of the global optimal layout scheme meets the evolutionary stagnation condition during the loop iteration. If it does, perform an adaptive immune clearing operation to break the optimization deadlock until the termination condition is met, and output the global optimal physical utilization rate and its corresponding board layout scheme.

[0110] During the iterative process of step S6, when the system detects that the globally optimal physical utilization rate has not improved for M (M=15) consecutive generations, it determines that evolution has stalled and triggers immune purge. The algorithm traverses the population to remove redundant individuals with immune concentrations higher than a preset threshold and injects an equal number of random new individuals. During the purge process, the system forcibly exempts the top E2 (E2=3) elite individuals and individuals containing dynamic vaccine sequences, ensuring that core superior genes are absolutely preserved while breaking deadlock.

[0111] Using the above-mentioned optimization algorithm, the optimal layout scheme found by the method of this invention is shown in Table 1. Here, I is the layout pattern number, II is the length of the solid wood board (mm), III is the length (mm) of the standard-sized boards arranged on the corresponding solid wood board (quantity / piece), IV is the remaining length of the solid wood board (mm), and V is the number of layout patterns / units. This scheme uses a total of 26 solid wood boards, achieving a board utilization rate of 99.41%. Furthermore, Table 2 shows the comparison results between the method of this invention and existing optimization algorithms.

[0112] Table 1. Optimal sorting schemes found by the method of this invention.

[0113]

[0114]

[0115] Table 2 Comparison of the present invention with other preferred methods

[0116]

[0117] To further verify the effectiveness and advancement of the method proposed in this invention, the following analysis combines the detailed layout scheme in Table 1, the comparative data in Table 2, and... Figure 2 The training process comparison curves are used to conduct an in-depth analysis of the beneficial effects of the technical solution in this application:

[0118] Based on the layout details in Table 1, this invention demonstrates extremely high spatial search capabilities. When solving large-scale short-board examples, this invention found a globally optimal solution with a physical utilization rate as high as 99.41%. In particular, layout patterns 7 and 22 achieved 0mm remaining length of the original board material, realizing 100% absolute full load, while most other patterns had remaining material within a very small range of 5-20mm. This proves that the forward-looking dynamic decoding and dynamic vaccine mechanism of this invention can effectively inherit excellent compact layout characteristics, approaching the physical limits of material utilization.

[0119] In addition, combining the data in Table 2 with Figure 2 The convergence curve of this invention shows that it significantly outperforms existing technologies in terms of escaping local optima and improving solution accuracy. For example... Figure 2 As shown, during the training process, traditional genetic algorithms stagnate in the early stages of iteration (around generation 5), exhibiting severe premature convergence; the polar bear algorithm gets stuck in a local optimum around generation 20; and the gray wolf algorithm shows violent oscillations. In contrast, the method of this invention (orange curve) shows a significant step-like upward trend, achieving breakthroughs even when encountering evolutionary stagnation multiple times. This indicates that by introducing immune purge and elite tabu search fine-tuning, the algorithm successfully breaks through population homogeneity and local extremum traps, ultimately achieving an average utilization rate of 99.41% (as shown in Table 2), an improvement of 0.22% compared to traditional algorithms.

[0120] The above description is only a preferred embodiment of the present invention. It should be noted that for those skilled in the art, several improvements and modifications can be made without departing from the principle of the present invention, and these improvements and modifications should also be considered within the scope of protection of the present invention.

Claims

1. A method for optimizing plate layout based on the fusion of immune genetics and taboo search, characterized in that, Includes the following steps: S1. Collect target board demand data and original board inventory data for solid wood boards, assign serial numbers to the target boards, and initialize the population using a random shuffling strategy to obtain multiple chromosome individuals. S2. Perform decoding operation on each chromosome individual in the population, allocate the target board to the original board according to the gene sequence, and calculate the smooth fitness and physical utilization rate of each individual for the allocated layout. At the same time, update the global optimal layout plan record based on the physical utilization rate. S3. Determine whether the current iteration count meets the preset taboo search cycle threshold. If it does, select the top E1 elite individuals in the population with the highest physical utilization rate and input them into the taboo search module to perform local depth fine-tuning. Then, replace the original elite individuals with the fine-tuned individuals. If not satisfied, proceed to S4; S4. Calculate the Hamming distance between individuals within the population to assess gene similarity, and calculate the immune concentration of each individual based on gene similarity; S5. Combine smooth fitness and immune concentration to calculate the overall selection probability of an individual, and perform genetic evolution operations based on immune selection according to the probability to generate a new generation of population. S6. Repeat S2 to S5, and monitor in real time whether the physical utilization rate of the global optimal layout scheme meets the evolutionary stagnation condition during the loop iteration. If it does, perform an adaptive immune clearing operation to break the optimization deadlock until the termination condition is met, and output the global optimal physical utilization rate and its corresponding board layout scheme.

2. The plate layout optimization method based on the fusion of immune genetics and taboo search according to claim 1, characterized in that, In S2, the decoding operation is performed as follows: The decoding operation adopts a dynamic optimal adaptation rule, which is as follows: the target board is read sequentially according to the gene sequence, and it is preferentially allocated to the opened original board with the smallest remaining space after allocation; if all the currently opened original boards cannot accommodate the target board, the subsequent target board sequence is calculated in advance, and a new specification original board with the highest expected filling rate is opened from the inventory.

3. The plate layout optimization method based on the fusion of immune genetics and taboo search according to claim 1, characterized in that, In S2, the smooth fitness and physical utilization rate of each individual are calculated, and the implementation content of the globally optimal sorting scheme record is updated based on the physical utilization rate: Calculate the physical utilization rate: Extract the total length of all target boards in the current individual layout plan, and the total length of the original boards actually consumed after allocation. Calculate the global physical utilization rate U of the layout plan by the ratio of the two. The calculation formula is as follows: ； in, This indicates the current physical utilization rate of an individual; Indicates the total number of target sheet materials; Indicates the first The length of the target sheet material; This indicates the total number of original boards used in this individual layout plan; Indicates the first The original material has been opened to its specified length; Calculate the smoothness fitness: Sum the target board lengths allocated to each opened original board, and calculate the single-board fill rate for each original board. Then, average the squares of the single-board fill rates for all opened original boards to obtain the individual smoothness fitness. To enhance the fitness weight of compact layout combinations, the calculation formula is as follows: ； in, This represents the smooth fitness of the current chromosome individual; This indicates the total number of original boards used in this individual layout plan; Indicates allocation to the first The number of target boards in the original board material; Indicates the first The first in the root board The length of the target plate.

4. The plate layout optimization method based on the fusion of immune genetics and taboo search according to claim 1, characterized in that, In S3, the implementation of local depth fine-tuning by the tabu search module is as follows: Multiple neighborhood solutions are generated using a two-point position swap strategy. If the physical utilization rate of a neighborhood solution is higher than the historical global best record, the contempt criterion is triggered, and the neighborhood solution is directly accepted. If the contempt criterion is not triggered and the move is not in the taboo list, the neighborhood solution with the highest fitness is accepted, and the move is added to the fixed-length taboo list.

5. The plate layout optimization method based on the fusion of immune genetics and taboo search according to claim 1, characterized in that, In S4, the specific implementation details are as follows: S401. Constructing the feature matrix: Transform the population into a matrix of size... The feature matrix; in, This represents the total number of individuals in the population. Indicates the total number of target sheet materials; Indicates the first The first individual The target plate number corresponding to each gene locus; S402. Calculate Hamming distance: Compare any two individuals in the feature matrix one by one. and By analyzing gene sequences and counting the number of genes with different values ​​at the same gene locus, the Hamming distance can be obtained. The calculation formula is as follows: ； in, Represents an individual With individuals Hamming distance between them; Indicates characteristic functions; S403. Calculate gene similarity: Assess gene similarity between individuals based on the calculated Hamming distance. The calculation formula is as follows: ； in, Represents an individual With individuals Genetic similarity between them; S404. Calculate the individual's immune concentration. The calculation formula is as follows: ； in, Indicates the first The immune concentration of an individual; This indicates the preset threshold for determining gene similarity; ( ) indicates an indicator function.

6. The plate layout optimization method based on the fusion of immune genetics and taboo search according to claim 1, characterized in that, In S5, the specific implementation details are as follows: The overall selection probability adopts an exponential decay mechanism based on concentration suppression, and the specific calculation method is as follows: ； in, Indicates the first in the population The overall probability of selection for each individual; , Indicates the first The, the The smoothness of fitness of each individual; , Indicates the first The, the The immune concentration of an individual; This represents the preset concentration inhibition coefficient; This represents an exponential function with the natural constant e as its base. It represents the total number of individuals in a population.

7. The plate layout optimization method based on the fusion of immune genetics and taboo search according to claim 1, characterized in that, In S5, the implementation of genetic evolution operations is as follows: The genetic evolution operation is as follows: a gene sequence from a parent chromosome is randomly selected and directly copied to the same position in the offspring. Then, a gene from another parent chromosome that does not appear in the above sequence is filled into the remaining empty space in the offspring in the original order to ensure the absolute conservation of the types and quantities of the target material for the short board.

8. The plate layout optimization method based on the fusion of immune genetics and taboo search according to claim 1, characterized in that, In S6, the specific implementation details are as follows: The determination of whether the evolutionary stagnation condition is met and the execution of the adaptive immune purge operation are as follows: when the physical utilization rate of the globally optimal sorting scheme has not improved for M consecutive generations, the algorithm is determined to be trapped in a local extreme value trap and meets the evolutionary stagnation condition, and the adaptive immune purge operation is triggered. The adaptive immune clearance operation specifically includes: traversing the current population, identifying individuals with immune concentrations higher than a preset clearance threshold as redundant antibodies and removing them, and then using a random initialization strategy to generate an equal number of new chromosome individuals to fill the population vacancies; and during the clearance process, exemptions are reserved for elite individuals with the highest physical utilization rate (E2) and individuals containing dynamic vaccine sequences.

Images